In fact, I believe that library-sector developers have asked Amazon and
been told that their use is allowed. But definitely, there's no
guarantee this will always continue be true. The terms of use don't seem
to have substantially changed to me, but they could always start
enforcing them more strictly -- for new accounts created to use the
Product Advertising API, it looks like there actually will be a manual
review step where Amazon staff approves you or doesn't, which never
existed before.
So, while I'm still using it, I'm also keeping in mind what backup plans
I have if they ever ask me to stop.
Here are the things I use Amazon API for, with alternates:
1) To take an ISBN, and look up more complete metadata for it.
Alternatives:
A) Google Books Data API (free for everyone; yes, there is a GBS API
which is explicitly authorized for non-javascript access. GBS API will
also allow you to find OCLCnums and LCCNs that correspond to an ISBN,
when GBS has that data, which it often does thanks to the OCLC
relationship.)
B) WorldCat API (OCLC members)
C) Books In Print API, although BiP seems to be making up their mind
about whether they'll throw this in for free with an existing BiP online
subscription, or charge extra for it.
D) OpenLibrary? (Is this true?)
2) Cover images. Alternatives:
A) CoverThing
B) Google Books
C) OpenLibrary
3) To find an ASIN, in order to make a link to the Amazon page.
Ironically, this is actually what the API is _for_, and what Amazon
would actually WANT you to do, but it's the thing that's least
replaceable. If you have the ISBN, and if you assume the ASIN is the
same as the ISBN, you don't need an API. This is often true, but not
guaranteed to be true, and I think will become less true when the new
ISBN-13 namespace starts to be used. In my case, I use the ASIN to
identify if Amazon has a search-inside and/or limited-excerpts
available, but the API actually doesn't support that, I've been
screen-scraping all along for that, once I have the ASIN.
Tim Spalding wrote:
They're also tightened up the API in various ways, and renamed it the
"Amazon.com Product Advertising API." Although I know of no case when
Amazon has shut down a library, it would be hard for any to claim
their site had "as their principal purpose advertising and marketing
the Amazon Site and driving sales of products and services on the
Amazon Site."
I think it's a terrible mistake for them. Their marginal cost is zero;
they don't need to do this. Data openness was a key factor in Amazon's
rise. And that was when thee were no other options. With viable other
options just emerging—Open Library, Google, at least—now is hardly the
time to make it less attractive.
Tim
On Mon, May 11, 2009 at 9:40 AM, Jonathan Rochkind <rochk...@jhu.edu> wrote:
The Amazon products API keeps changing it's name, and has just been changed
to Amazon "Product Advertising API" -- it's the one you use to look up books
in Amazon and get metadata for them, though.
It looks from an email I got from Amazon that ss of August 15th, you'll need
to cryptographically sign requests to this API, to have them responded to.
It looks like kind of a pain.
I think a bunch of people on this list may be using this API. Beware.
Instructions for how to cryptographically sign requests the way they want
can be found here:
http://docs.amazonwebservices.com/AWSECommerceService/latest/DG/Query_QueryAuth.html
http://docs.amazonwebservices.com/AWSECommerceService/latest/DG/rest-signature.html
Like I said, it's looking like a pain to me. There are lots of details to
get right. If you URI-escape not _exactly_ the same way they do, it's not
going to work. Etc.
Jonathan
